In Helfrich, H. (Ed.), Time and mind (pp. 171-195). ;J Kirkland, WA: Hogrefe & Huber, 1996. :l Chapter 9: Models of psychological time revisited* RICHARD A. BLOCK and DAN ZAKAY Introduction Theorists have taken two seemingly different approaches to ex­ plaining, or modeling, psychological time (Block, 1990). These ap­ proaches have appeared under several guises. Ornstein (1969) referred to them as tlle sensory-process approach and the cognitive approach. Sensory-process models "postulate some sort of <time-base', a repetitive, cumulative, pUlse-dispensing mechanism which delivers internal time signals, an 'organ' of time" (p. 25). Ornstein claimed that this type of model has not provided a useful way to understand duration experience. This approach may also have difficulties explaining why cognitive, or infonnation-processing, variables influence duration experience. TI1e other class of model includes various' proposals concerning the important cognitive factor underlying duration experience, such as "images" (Guyau, 1890~ 1988), "changes," (Fraisse, 1957; 1963), "mentaI con­ tent" (Frankenhaellser, 1959), "storage size" (Ornstein, 1969), and "contextual changes" (Block & Reed, 1978). According to some propo­ nents of sensory-process models, cognitive models cannot easily explain tl1e near-linear psychophysical relationship between physical and psy­ chological duration, as well as the possible influence of physiological variables sllch as body temperature. Timing ,vith or \vithout a timer TIle important difference between tlle two approaches is not that the first concerns sensory processes and tllat the second concerns cognitive processes. Instead, the first class of model proposes timing witll a timer, whereas the second proposes timing WitilOllt a timer (Ivry & Hazeltine, 1992). In timing-with-a-timer models, a pacemaker mechanism lUlderlies • We thank Hannes Eisler, Fran90ise Macar, John Moore, and Andras Semjen for helpful comments on an earlier draft of this chapter. , 1/_1 l72 Richard A. Block and Dan Zakay :1 Models of psychological lime revIsIted .~'~l the psychological timing system. Two major variants are chronobiologi­ ~ more general. Consider, for example, Doob's (1971) model (Figure 1, p. 174). This model illustrates a taxonomy of time tllat depicts interactions cal and intemal-clock models (Block. 1990). J In timing-without-a-timer models, subjects construct psychological time from processed and stored 1:1' involving the "intricate. multivariate phenomenon of time" (p. 30). The information - that is, some salient aspect or byproduct of infonnation '~I details are relatively unimportant. For present purposes, we note tllat processing. Valiants of this kind of model include attentional, memory ~ altll0ugh tllis model may seem comprehensive, it is not a fW1ctional (e.g., :( infon1lation-processing) model of temporal behavior or judgment. storage, and memory change nlodels (Block, 1990). .~. ~ Pieron (1923) was one of the first researchers to discuss the possi­ ~ Block (1985) proposed a contextualistic model in which temporal ble relationship between body temperature and duration experience. i experience is a product of four kinds of interacting factors (Figure 2, p. 175). Again, tlle details are relatively unimportant, because model is Subsequently, Frant;ois (1927) and Hoagland (1933) obtained such evi­ ~r tllis dence. which SUppOlts a possible timing-with-a-timer model. Hoagland only a little more functionalistic than Doob's. 111e main advantage of f~ (1933; see also 1966) proposed that a masler chemical clock, or tempo­ R models such as Doob's (1971) and Block's (1985) is heuristic: tllese ral pacemaker, in the brain regulates time-related behaviors and judg­ ~ models remind us that psychological time involves complex interactions ments. TIle evidence suggested that the rate of repeated time productions , of various organismic and environmental variables. 111e main disadvan­ - involving counting at the rate of one per second - increases as a func­ , tage of these models, as noted ab~ve, is that they do not "relate in a fW1ctional way to the empirical findings [they are] supposed to repre­ tion of body temperature. More recent evidence suggests that duration ~ f sent" (Michon. 1985, p. 26). Although both models depict interactions judgments of many minutes (e.g., hourly productions) are correlated "~ with body temperature (Campbell & Bimbaum, 1994). Although the re­ of variables, several functional issues remain: (a) Which interactions are lationship between body temperature and shorter duration judgments is impOItant in particular situations and which are not? (b) What is tlle often inconsistent (Hancock, 1993), changes in body temperature do nature ofthe higher-order interactions? (c) How are the tmderlying proc­ seem to lead to systematic changes in the rate of psychological time esses sequenced, as in a functionalistically oriented infonnation-process­ (Wearden & Penton-Voak, 1995). One possibility is that body tempera­ ing model of temporal beha vior? I, ture influences general arousal level, which thereby influences the rate of i~ ;~ Cognitive psychologists and others have occasionally proposed a pacemaker mechanism (Wearden & Penton-Voak, 1995). The problem models resembling internal-clock models, but tllese usually involve tim­ with postulating that a pacemaker or master biochemical clock directly ing without a timer. For example, Lashley (1951) tllOught tllat practiced influences time-related behaviors and judgments is tllat temperature may movement sequence,s are structured as individual elements organized into also influence brain processes tllat subserve attentional, memory, and chunks which are executed as part of a motor program for tlle action se­ otller cognitive processes. Variations in tllese processes probably have quence. Because he proposed tllat a motor program is executed without little or no effect on body temperature. Because cognitive variables (e.g., the need for feedback, it needs an internal-control process to time ele­ attentional demands of a task) influence duration experience. cognitive ments. Researchers have searched for such a common mechanism tllat is processes may directly mediate temporal behaviors and judgments. Body able to stabilize motor programs despite changes in states of the organ­ temperature may indirectly influence temporal behaviors and judgmehts ism, changes in contextual stimuli, changes in equipment or instruments by altering whatever cognitive processes subserve psychological time used for the pClfonllance, and so on. 111e important question of how (Block, 1990). movement sequences are timed is still largely lmresolved, as is the ques­ Theorists have proposed a large number of cognitive models of tion of whether we need to propose an intemal-c1ock mechanism. Motor programs may contain internal, hierarchically organized infonnation psychological time. TIley have stated these mostly In tlle form of a "variable-x hypothesis:' where one may substitute any of several vari­ about timing relationships, so a pacemaker mechanism may be lumeces­ ables for variable-x (e.g., input segmentation, complexity-of-coding, at­ sary. Altemativeiy, even sllch information about timing relationships tentional selectivity). Each of these variables is typically tlle only one may rely on a pacemaker for some basic calibration (see Semjen, this that the researcher manipulated. A few models have attempted to be volume, chapter 2). :~ An example of an internal-clock model is also desribcd in chapter 4 by H. 'If i Fi~l('r. ;i 174 Richard A. Bloch. ami Dan Zakay \ Models of psycho)ogicLd lillie revisited 175 .114 In the remainder of the chapter, we review various fonnal models of :~ psychological time. We propose the attentional-gate model, which recon­ :q ciles the two approaches. This model is somewhat isomorphic with ·iit ;il contextual-change models of experienced and remembered duration. ::.1 ~i IJ CHARACTERISTICS J OF EXPERIENCER .:\ '~ Species Sex Personality ] Interests 1 Previous experiences 1 1 CONTENTS OF ACTIVITIES DURING I TIME PERIOD(S) TIME PERIOD(S) 1 "Empty" 'Passive" nonattendlng "Filled' 'Passive" attending Active responding Active responding (linguistic (level of processing. pictorial. kind of encoding, I music etc.) strategies, Complexity etc.) TEMPORAL BEHAVIOR (METHODS) I i Judgments/Estimates of: simultaneity successiveness rhythm serial position order I 1 I spacing duration 1 Figure 2: Block's (1985) contextualistic model ot duration expenence. Treislllan's Model Treisman (1963) proposed an influential model of an internal clock underlying human temporal judgment (Figure 3, p. 176). He postulated a pacemaker that produces a regular series of pulses, the rate of which varies as a ftmction of input from an organism'S specific arousal center. In his view, specific arousal is influenced by external events, in contrast to general arousal. which depends on internal mechanisms such as those tmderlying circadian rhythms. A counter records the number of pulses in a pathway, and the total is transferred into a store and into a comparator mechanism. A verbal selective mechanism assists in retrieving useful in­ fom1ation from the store. 111i5 is presumably a long-tenn memory store Figure I: Doob's (1971) taxonomy of time \7ti Richard A. Ulud. LlUJ Uau Lakay (.IUll;,:.!,; 01 p.})~I!I')I"i..>ILdi LllI ..... 1,_, containing knowledge of correspondences between total pulses and ver­ pulse rate. In our view, this minor modification docs not rectify the ballabels, sllch as